Notes on the challenges of regulating technology businesses
Professor Diane Coyle spoke at a recent Trust & Technology Initiative lunchtime event, about regulating the tech industry. It was great to get a thoughtful outline of some key issues which so rarely get discussed in the excitement over what to do about Big Tech, from someone who actually has deep experience of the economics and regulatory space.
I'm writing these overdue notes about Diane's talk on the day when the Lords Communications Committee published its report on regulating in a digital world. It's nice that some of Doteveryone's ideas for an Office of Responsible Technology to help empower regulators seem to have been picked up in the report. Still, Heather Burns's thoughts seem particularly pertinent:
Tech is more than "big tech." Some of it is made in the UK (and Heather's afterbrexit.tech is a super resource for working out what might change for UK tech...). Anyway.
Furthermore, to regulate an industry requires a legally-robust definition of a group of companies. this is hard to achieve in 'tech'. You also need a jurisdiction willing to take action, or international coordination. There's an opportunity cost - regulators are uncomfortable with predicting the future. We should also remember that regulation is a burden on other companies, even if we are mostly concerned with Facebook, Google and so on.
In tech, the usual tools for competitive analysis don't work. There's the two-sided market aspect; profits are hard to analyse; and 'envelopment' where a consumer base is built in one area and then moved to another market. Competition policy is generally legalistic, but lawyers use different types of evidence to economists, so it's harder to agree things.
Another challenge to actually doing regulation is that big corporates tend to work with (or buy up) all the professional consultancies in town, so they all end up being "conflicted out" of any action. (This perspective was totally new to me - and sort of obvious for corporates with large budgets for lobbying and other activities.)
There are some more unusual regulation ideas being implemented around the world. Germany's recent decision on Facebook data linking, for instance, or India's ruling that Amazon cannot sell its own products on its platform.
As well as monopoly issues, there are consumer protection ones, where powers are often already available - around unreasonable terms and conditions, harms to children, and harms from advertising.
The ecosystem matters as well as individual cases. Competition authorities need to be able to look at mergers collectively, not individually. Diane noted that the cost of advertising for "sharing economy" services (Uber, AirBnb, etc) is going up on Google and Facebook.
The consumer value for free services is a really tough nut to crack though. It's hard to act when consumers feel they are getting a good deal. A comparison might be made with free current account banking - we are all used to this in the UK, and it's very hard for a government to say you now have to pay for it. So what other methods might we try? We could look at requiring data sharing and interoperability; for instance, trying to make social media a bit more like email where there's a common interchange standard. Or we could insist on terms of access to data to reduce the advantage of holding huge datasets. Algorithmic collusion in pricing could be made an offence - there is no proof that this is happening, although it is suspected.
Any action though requires political will, and our political leaders want tech business here (such as the huge Google site at Kings Cross in London). With Brexit and the state of the economy in general, pushing back on big tech is not appealing. It is not clear if there would be an appetite for betting on new entrants to the market if Google and Facebook were diminished here.
So there needs to be a cross-cutting regulatory structure and an international framework, which would look at ads, personal data, and other issues together. The value of data needs to be understood better, as data has weird characteristics, beyond that it is non-rival. Sometimes it may depreciate over time - or not. The value exchanges can only be understood and reasoned about if we have greater understanding here (and Diane and the Open Data Institute are looking into this - last year's workshop was a great start). We also need to consider the incredible value of less-debated services, such as mapping combined with GPS. Issues of data, though, feel easy for regulators to approach, compared to the end to end services such as Amazon's logistics platforms.
Data is not the new oil or the new gold or even the new asbestos. Perhaps we should think of "data national parks" - assets to be looked after, stewarded and value realised for society overall.
In the Q&A, the usual question of "can regulators keep up with the speed of tech" came up. Diane suggested that principles-based regulation was the way to address this, perhaps backed up by a "flying squad" of tech-savvy people (public interest technologists!) who could be borrowed by regulators on demand. We should also remember that new technologies and disruptors do come along. MySpace was big, once. Hal Varian, chief economist at Google, is concerned that the next wave of tech might appear whilst 'big tech' management is distracted by today's regulation discussions, and that today's big companies might be disrupted and supplanted.
We need our regulators to have courage in the tech space, just as in the finance sector.
I'm writing these overdue notes about Diane's talk on the day when the Lords Communications Committee published its report on regulating in a digital world. It's nice that some of Doteveryone's ideas for an Office of Responsible Technology to help empower regulators seem to have been picked up in the report. Still, Heather Burns's thoughts seem particularly pertinent:
 |
| https://twitter.com/WebDevLaw/status/1104307527095894016 |
Tech is a tough area to regulate, for various reasons. Firstly you have both strong network effects, acting on both sides of two-sided markets, so it's doubly tricky. There's a winner takes all dynamic based on the scale of the loss that can be sustained by a business, and in Silicon Valley with plentiful capital that loss can be very large indeed.
Small countries have limited options for action, so everything comes down to what big countries (US) or blocs (EU) do.
Another challenge is that the consumer gets incredible value from 'big tech'. Research (looking at how much you'd need to be paid to give up an internet service) suggested that online search alone was worth $17.5k per year to a consumer (and that's in the US, where the average income is $31k). Even with the costs we each bear, of network access, buying devices (phones) etc, and even advertising value, the benefits are greater. The socio-economic aspects are interesting too; in the UK, it seems that consumers value these services more highly in areas with lower incomes, which is odd.
Furthermore, to regulate an industry requires a legally-robust definition of a group of companies. this is hard to achieve in 'tech'. You also need a jurisdiction willing to take action, or international coordination. There's an opportunity cost - regulators are uncomfortable with predicting the future. We should also remember that regulation is a burden on other companies, even if we are mostly concerned with Facebook, Google and so on.
In tech, the usual tools for competitive analysis don't work. There's the two-sided market aspect; profits are hard to analyse; and 'envelopment' where a consumer base is built in one area and then moved to another market. Competition policy is generally legalistic, but lawyers use different types of evidence to economists, so it's harder to agree things.
Another challenge to actually doing regulation is that big corporates tend to work with (or buy up) all the professional consultancies in town, so they all end up being "conflicted out" of any action. (This perspective was totally new to me - and sort of obvious for corporates with large budgets for lobbying and other activities.)
There are some more unusual regulation ideas being implemented around the world. Germany's recent decision on Facebook data linking, for instance, or India's ruling that Amazon cannot sell its own products on its platform.
As well as monopoly issues, there are consumer protection ones, where powers are often already available - around unreasonable terms and conditions, harms to children, and harms from advertising.
The ecosystem matters as well as individual cases. Competition authorities need to be able to look at mergers collectively, not individually. Diane noted that the cost of advertising for "sharing economy" services (Uber, AirBnb, etc) is going up on Google and Facebook.
The consumer value for free services is a really tough nut to crack though. It's hard to act when consumers feel they are getting a good deal. A comparison might be made with free current account banking - we are all used to this in the UK, and it's very hard for a government to say you now have to pay for it. So what other methods might we try? We could look at requiring data sharing and interoperability; for instance, trying to make social media a bit more like email where there's a common interchange standard. Or we could insist on terms of access to data to reduce the advantage of holding huge datasets. Algorithmic collusion in pricing could be made an offence - there is no proof that this is happening, although it is suspected.
Any action though requires political will, and our political leaders want tech business here (such as the huge Google site at Kings Cross in London). With Brexit and the state of the economy in general, pushing back on big tech is not appealing. It is not clear if there would be an appetite for betting on new entrants to the market if Google and Facebook were diminished here.
So there needs to be a cross-cutting regulatory structure and an international framework, which would look at ads, personal data, and other issues together. The value of data needs to be understood better, as data has weird characteristics, beyond that it is non-rival. Sometimes it may depreciate over time - or not. The value exchanges can only be understood and reasoned about if we have greater understanding here (and Diane and the Open Data Institute are looking into this - last year's workshop was a great start). We also need to consider the incredible value of less-debated services, such as mapping combined with GPS. Issues of data, though, feel easy for regulators to approach, compared to the end to end services such as Amazon's logistics platforms.
Data is not the new oil or the new gold or even the new asbestos. Perhaps we should think of "data national parks" - assets to be looked after, stewarded and value realised for society overall.
In the Q&A, the usual question of "can regulators keep up with the speed of tech" came up. Diane suggested that principles-based regulation was the way to address this, perhaps backed up by a "flying squad" of tech-savvy people (public interest technologists!) who could be borrowed by regulators on demand. We should also remember that new technologies and disruptors do come along. MySpace was big, once. Hal Varian, chief economist at Google, is concerned that the next wave of tech might appear whilst 'big tech' management is distracted by today's regulation discussions, and that today's big companies might be disrupted and supplanted.
We need our regulators to have courage in the tech space, just as in the finance sector.